Replication package for Zhuang, Maiting. 2022. “Intergovernmental Conflict and Censorship: Evidence from China’s Anti-Corruption Campaign”. Journal of the European Economic Association

Replication package structure:

From root directory (set your root directory by changing the line "global dir "your directory path"" at the beginning of each do-file)
1. data (contains all raw data required for replication, all intermediate datasets saved here)
2. code (contains all code required for replication)
3. output (empty folder, all figures and tables saved here)


List of files:

1. data:

Raw data included in folder:

officials.dta - data on officials under investigation
newspapers.dta - data on newspapers in WiseNews sample
officialnewspaper.dta - matrix of all officials and newspapers
of_articles.dta - data on articles (at official-article level)
of_prov_code.dta - code for official province
np_prov_code.dta - code for newspaper province
province_covariates.data - data on provincial level covariates
baidu.dta - province-day level data on Baidu searches for individual officials and CCDI complaint website
baidu_reg.dta - data on Baidu searches for officials for year after investigation announcement
sinaweibo.dta - panel of Sinaweibo comments by province of SinaWeibo commenter
sinaweibo_comments.dta - data on Sinaweibo comments

Datasets constructed using Datasets.do:

articles.dta - data on articles
ofnppanel.dta - official-newspaper-day panel
ofnppanel_week.dta - official-newspaper-week panel
ofnppanel_month.dta - official-newspaper-month panel
chartdata.dta - data on mean article by own-province, high-level official and post-investigation
ofnppost.dta - data on mean article by newspaper, official and post-investigation
np_by_date.dta - data on newspaper by date
of_timeline.dta - panel data on official investigation announcements
np_timeline.dta - panel data on number of provincial newspapers in sample
art_timeline.dta - panel data on characteristics of articles about investigated officials

Datasets constructed using Appendix A.do:

npchart_sub.dta - data on matched subsidiary-owner newspaper pairs


Datasets constructed using Appendix B.do:

prov_coefs.dta - province coefficients


2. code

Datasets.do - constructs additional datasets used in the analysis
holidays.do - dates for Chinese holidays and CCP events
Main Text - Panel Regressions.do - Tables and figures based on ofnppanel in main text. Table 2, Figures 2-6 (and Appendix Tables E1-E5)
Main Text - Other.do - Other tables and figures in main text. Tables 1, 3-5, Figure 1
Appendix A.do - Tables and figures for Appendix A (for Figures A2 and A5, need to run code for Table B2 first for prov_coefs.dta)
Appendix B.do - Tables and figures for Appendix B
Appendix C.do - Tables and figures for Appendix C
Appendix D.do - Tables and figures for Appendix D


Software requirements:

Stata version MP/15.1 4-core

Stata packages: 
reghdfe (version 5.7.3 13nov2019)
ftools (version 2.37.0 16aug2019)
ppmlhdfe (version 2.2.0 02aug2019)
egenmore
areg
estout
coefplot
regsave

Please note that ofnppanel.dta contains 39,941,978 observations and its construction requires large computing memory.


